Temperature Change in the EU

by Maximilian Mrachacz
for IronHack Remote Bootcamp Data Analytics

Overview:

Finding parameters that (might) have an impact on / drive Temperature Change.
Final Version: New_Version_Temp_Change_and_other

Tools and Methods Used:

- Python               
- Statistical analysis             
- Data visualization              
- Jupyter Notebook              
- Tableau             
- Machine Learning (Logistic Regression)                
- About the Dataset:              
- Data is about Temperature Change and Factors that might be correlated.                

Sources:

- https://www.fao.org/faostat/en/#data               
- http://data.un.org/Explorer.aspx             
- Final Dataset/DataFrame:              

About New_Version_Temp_Change_and_other.ipynb:

- 838 rows × 14 columns                 
- created by merging 10 sub- tables together             
- columns hold exactly the information as what there called              
- last two (predicted) columns come from prediction with GradientBoostingRegressor              
- tried different models, compared precision, decided on used model          
- for prediction, 2 new dataset were created with manufactured data values     

Data Cleaning:

- dropping columns       
- converting numbers         
- merging, renaming           
- filling NaNs (mostly by KNeighbor Regression)          
- using 3 neighbors     

Data Storage:

- Saved data as .csv file               

Visualization: Tableau

Presentation: Prezi

Conclusion:

- The Parameters that were examined suggest correlations with the rise of average temperatures. 
- For better prediction result, more Parameters are needed